620

Langmuir 1986,2, 620-625

Solid Surface Tension: The Interpretation of Contact

Angles by the Equation of State Approach and the Theory
of Surface Tension Components
J. K. Spelt,t D. R. Absolom,tJ,and A. W. Neumann*tJl
Department of Mechanical Engineering, University of Toronto, Toronto, Ontario M5S 1A4,
Canada, Institute of Biomedical Engineering, University of Toronto, Toronto,
Ontario M5S l A 4 , Canada, and Research Institute, Hospital for Sick Children, Toronto,
Ontario M5G 1x8,Canada
Received August 7, 1985. I n Final Form: May 28, 1986
The two methods which are frequently used to determine solid surface tensions, the equation of state
approach and the theory of surface tension components, often yield conflicting results and represent two
completely different ways of conceptualizing interfacial tension problems. In this report an experiment
is described which permits the independent evaluation of these two theories as part of an effort to establish
the relative merits of each. The contact angles of two different liquids on a single solid substrate are found
to be identical when the total liquid surface tensions are equal. This result, which was obtained for five
liquid pairs on different surfaces, is independent of the relative magnitudes of the dispersion forces within
the liquids of a given pair. This fiiding implies a basic deficiency of the theory of surface tension components
and supports the equation of state approach.

Introduction
The determination of solid and solid-liquid surface
tensions is of importance in a wide range of problems in
pure and applied science. Since it is not possible to
measure directly surface tensions involving a solid phase,
there exist, a t present, many indirect approaches for obtaining these values. These various methods are often in
considerable disagreement, both quantitatively and from
a theoretical standpoint. The problem persists because
most of these approaches have not been objectively tested
through the prediction of physical phenomena which could
be independently observed and thus used to validate the
various theories.
The purpose of the present investigation was to apply
one such independent test to the following two theories:
(1)the Fowkes theory of surface tension components;l (2)
the equation of state approach to interfacial tensions.2
These two approaches for evaluating solid surface tension
are briefly outlined below.
(1)The theory of surface tension components was pioneered by Fowkesl who proposed that surface tension
should be considered in terms of components, each due
to a particular kind of intermolecular force. Thus, a given
organic liquid may have discrete surface tension components attributable to London dispersion forces, dipoledipole (Keesom) forces, induction (Debye) forces, and
hydrogen-bonding forces. Such surface tension components, although not thermodynamically defined, are,
nevertheless, regarded by Fowkes as unique physical
properties of the material.' Liquid-solid interactions are
considered only between those surface tension components
which arise from the same types of forces. Therefore, at
a water-Teflon interface, for example, since only dispersion
forces are present in the Teflon, the large polar and hydrogen-bonding forces in the water will not act across the
interface to affect the interfacial tension directly. Thus,
in this approach the solid-liquid interfacial tension is a
function of the types and relative magnitudes of the intermolecular forces in the solid and the liquid.
The Fowkes equation and methdology is used to measure only the dispersion component of surface tension of
'Department of Mechanical Engineering,-University of Toronto.
Institute of Biomedical Engineering, University of Toronto.
8 Research Institute, Hospital for Sick Children.

0743-7463/86/2402-0620$01.50/0

a solid (or of a liquid). There are no widely accepted

methods for determining nondispersion components of
surface tension and, therefore, little consensus exists regarding the magnitudes of total solid surface tensions.
(2) The equation of state approach is based on thermodynamic arguments3 which lead to the conclusion that
the solid-liquid interfacial tension is only a function of the
total solid and liquid surface tensions. Unlike the Fowkes
approach, the types and relative magnitudes of the intermolecular forces in either phase are not considered to
be directly relevant. The interfacial tension is believed
to be completely defined by the total surface tensions of
the separate phases.
In the case of both of these theories, solid surface tensions are evaluated by contact angle measurements. From
Young's equation,
YSL = Ysv - YLV cos 0
(1)
it is seen that if yLvand ysv, the liquid-vapor and solidvapor surface tensions, respectively, are fixed for a series
of solids and liquids, then the contact angle is directly
related to ysL, the solid-liquid interfacial tension. This
concept forms the basis for a direct experimental comparison of the predictions of the two theories.
Consider first two different pure liquids that are chosen
to have equal overall surface tensions, as measured by, for
example, the Wilhelmy plate technique. These same liquids are, however, also selected to have widely disparate
compositions of intermolecular forces. In other words, one
liquid may be an alkane (a liquid which has only dispersion
forces) while another may be characterized by a large
dipole moment and perhaps significant hydrogen bonding.
According to the theory of surface tension components, the
contact angles of these two liquids on a single solid surface
should differ in proportion to the differences in the
make-up of the intermolecular forces. In contrast, the
equation of state approach predicts that the contact angles
be equal since both the total liquid and solid surface
tensions are constant. This simple experiment provides
a direct test of the basic premise of each of the two theories, and, moreover, it is independent of the specific form
(1) Fowkes, F. M. Znd. Eng. Chem. 1964, Dec, 40-52.
(2) Neumann, A. W.; Good, R. J.; Hope, C. J.; Sejpal, M.J. Colloid

Interface Sci. 1974, 49, 291.

(3)Ward, C.A.; Neumann, A. W. J. Colloid Interface Sci. 1974, 49,
286.

0 1986 American Chemical Society

Langmuir, Vol. 2, No. 5, 1986 621

Solid Surface Tension

Table I. Liquid Properties
dipole
moment:
surface tension,*
D
mJ/m2
liquid
pentadecane
0.0
28.93 - 0.085312'
0.97
42.14 - 0.10542'
dibenzylamine
0.23
41.82 - 0.11882'
1-methylnaphthalene
2.77
43.24 - 0.11952'
benzaldehyde
ethyl caprylate
1.68
29.12 - 0.10182'
2.58
28.50 - 0.07662'
heptaldehyde
2.23
41.84 - 0.12012'
methyl salicylate

6d/6T x
l O O C

100
71

79
60
50
47
44

Reference 18. *Measured by the Wilhelmy plate techniqueB

with an uncertainty of k0.15 mJ/m2. Fraction of total solubility
parameter ( 6 ~ attributed
)
to dispersion forces at 25 O C . Calculated
by using the solubility parameters in ref 5 and the correlations
suggested in ref 4 (for details see ref 9).

Table 11. Predictions of the Beerbower Correlation of

Liquid Surface Tension, yLV,with Solubility Parameter
Components (&, ,6, and tih). Comparison of Predicted and
Measured Surface Tensions (25 "C) and the Percentage of
the Total Predicted Surface Tension due to the Dispersion
Solubility Parameter Component (hd)
Beerbower measured Beerbower
YLV,
YLV,
YLV due to
liquid
mJ/m2
mJ/m2
6d, 70
pentadecane
29.6
26.8
100
dibenzylamine
41.0
39.5
94
I-methylnaphthalene
38.8
38.8
97
benzaldehyde
39.7
40.2
76
32.5
26.6
58
ethyl caprylate
heptaldehyde
31.4
26.6
56
methyl salicylate
39.2
38.8
56

Liquids. In order to interpret the significance of the final

contact angle results correctly, it is desirable to characterize each
liquid in terms of the relative magnitudes of dispersion and
nondispersion forces. In keeping with the goals of the experiment,
this should be done independently of the methodology of Fowkes.
A semiquantitative, relative assessment of the magnitudes of
nondispersion forces may be achieved either by comparing the
dipole moments or the empirical solubility parameters of the
various liquids.
The solubility parameter concept has been used successfully
as a practical tool in a wide variety of areas: Solubility between
two liquids is predicted on the basis of the degree of matching
encountered among three components of the solubility parameter;
viz., the dispersion (&), polar (8J, and hydrogen-bonding (&)
components. This is analogous to the familiar adage that "like
dissolves like", with the three empirically determined components
defining the molecular character of a liquid. Although such
solubility parameters lack a rigorous theoretical basis, they continue to be widely used in practice. Therefore, it is possible, with
some confidence, to assume that at least a semiquantitative
measure of the relative importance of dispersion forces can be
obtained through the consideration of the relative magnitudes
of published solubility parameters.
The first three liquids in Table I were selected to be significantly
more "dispersive" than the remaining four liquids, which are
characterized by much larger dipole moments and by relatively
smaller dispersion components of the solubility parameters. In
addition, the prediction of the Burrell "hydrogen-bonding"
classification: which is another empirical aid for the prediction
of solubility, is "moderate" for the last four liquids and "poor"
for the rest. This is not meant to imply the actual existence of
hydrogen bonding in our systems but in the present context serves
to indicate that independent experimental observation has established significant differences in the character of the intermolecular forces.
Taken together, the information in Table I indicates that,
relative to the last four liquids, dispersion forces in pentadecane,
dibenzylamine, and I-methylnaphthalene are responsible for a
significantly larger fraction of the total intermolecular binding
energy. For the purpose of the present investigation, the exact
magnitudes of such differences are unimportant. It is only
necessary to establish that there are appreciable differences in
the relative degree of "dispersion" in the first three liquids as
compared to the last four liquids.
BeerboweP has developed a correlation between liquid surface
tension and the dispersion, polar, and hydrogen-bonding components of the solubility parameter. Table I1 compares the

predictions of the Beerbower correlation with the measured surface

tensions for the seven liquids. The third column of this table lists
the percentage of the total predicted surface tension which is due
to the dispersion component of the solubility parameter. It should
be noted, as Beerbower himself did: that this dispersive fraction
bears no relation to the "dispersion component of surface tension"
as presented in the Fowkes theory. The purpose of Table I1 is
simply to provide another indication that pentadecane, dibenzylamine, and I-methylnaphthalene are characterized by a
substantially higher fraction of dispersion forces than are the other
four liquids.
In order to minimize the potential for vapor adsorption, the
seven liquids were chosen to have relatively high boiling points,
the lowest being that of heptaldehyde at 153 "C. The sources
of the liquids and the purities were as follows: benzaldehyde,
Fluka, Puriss grade; dibenzylamine, Aldrich, 99% grade; ethyl
caprylate (octanoic acid ethyl ester), Aldrich, 99+% grade, 99.9%
pure; heptaldehyde, Aldrich, 95% grade, 99.5% pure; 1methylnaphthalene and pentadecane, API Standard Reference
Materials (Carnegie Mellon University), 99.78 and 99.93 mol %,
respectively; methyl salicylate (2-hydroxybenzoic acid methyl
ester), Fluka, Purum grade, 99+%. The ethyl caprylate and
heptaldehyde commercial samples were further purified by
preparative gas chromatography prior to the contact angle experiments.
Solid Surfaces. The acquisition of thermodynamically significant contact angle data is largely dependent on the quality
of the substrate surface. The effects of roughness and heterogeneity can easily overshadow the influence of interfacial energ e t i c ~ . ~I t is, therefore, important in a study of this type to
produce solid surfaces of sufficient quality to ensure that the
observed contact angles accurately reflect the interaction between
the solid and liquid surface tensions as given by Young's equation,
eq 1.
Contact angle measurements were performed on two surfaces.
The first is heat-pressed Teflon FEP (Du Pont), a surface which
is exceptionally smooth and homogeneous. At 24 "C the surface
gave rise to advancing and receding contact angles of 5 2 O and 49",
respectively, for hexadecane, and 46" and 43", respectively, for
undecane. The method of preparation of this surface is summarized in ref 8, with greater detail supplied in ref 9.
In the present work, two F E P samples (designated A and B)
were employed, each having been prepared in a different way and
each having a unique thermal history. The latter fact caused the
surface tensions of the samples to be slightly different.
A second solid surface, which was used for only one pair of
liquids, was siliconized glass (details of dimethyldichlorosilane
treatment given in ref 9). The advancing contact angle with water
was 105", while the receding angle was between 95" and 100".
Contact Angle Measurements. Sessile-drop contact angles
were measured by a new technique (axisymmetric drop shape
analysis or ADSA) which fits the Laplace equation of capillarity
to an arbitrary m a y of coordinate points selected from the profiie
of a drop."l' This approach is unique because it does not depend

Barton, A. F. M. Chem. Reu. 1975, 75, 731.

Barton, A. F. M. CRC Handbook of Solubility Parameters and
Other Cohesion Parameters; CRC Press: Boca Raton, FL, 1983.
(6) Beerbower, A. J. Colloid Interface Sci. 1971, 35, 126.

(7) Neumann, A. W. Adu. Colloid Interface Sci. 1974, 4, 105.

(8) Spelt, J. K.; Rotenberg, Y.; Absolom, D. R.; Neumann, A. W.
Colloids Surf., in press.
(9) Spelt, J. K. Ph.D. Thesis, University of Toronto, 1985.

of any Fowkes-type equation or of any particular equation

of state.
The materials, methods, and results of this experiment
are outlined below.

Experimental Section

(4)
(5)

622 Langmuir, Vol. 2, No. 5, 1986

Spelt et al.

Table 111. Grand Average Advancing Contact Angles of All Available Experments on Substrate FEP A
grand avb
errorC
contact angled
exptl
liquid surface
no. of
liquids"
contact angle
limits (i)
difference
temp, OC
tension, mJ/m2
expts/drops
24
39.0
1.4
M
72.6
-0.2
24
39.0
0.7
MS
72.8
41.8
3
0.6
75.4
D
42.9
+2.0
3
0.4
73.4
B
3
41.8
0.6
75.4
D
41.5
+2.5
3
0.6
72.9
MS
25.6
39
0.7
52.4
P
25.5
-0.8
39
0.3
53.2
H
14
27.7
0.3
53.6
P
+0.6
14
27.7
0.4
53.0
EC

B, benzaldehyde; D, dibenzylamine (predominantly dispersive); EC, ethyl caprylate; H, heptaldehyde; M, 1-methylnaphthalene (predominantly dispersive); MS, methyl salicylate; P, pentadecane (totally dispersive). * Grand average contact angle: the average of the mean
contact angles (the average contact angle for a single experiment) for the number of experiments indicated. CErrorlimits: The "worst
possible case" values required to encompass all mean contact angles from the grand average contact angle. In cases where only one experiment was performed, the 95% student-t confidence limits on the mean contact angle are quoted. dContact angle difference: The contact
angle of the predominantly dispersive liquid (the first one listed in each pair) minus that of the significantly nondispersive liquid.
Table IV. Grand Average Advancing Contact Angles of All Available Experiments on Substrate FEP B
grand avb
errorC
contact angled
exptl
liquid surface
no. of
liquids"
contact angle
limits (i)
difference
temp, OC
tension, mJ/m2
expts/drops
3
41.8
4/12
D
72.4
3.2
0.6
+3.1
3
41.5
4/12
MS
69.3
39
25.6
1/3
P
49.1
0.5
H
50.3
0.4
-1.2
39
25.5
113
n-d

See footnotes of Table 111.

Table V. Grand Average Advancing Contact Angles of All Available Experiments on the Siliconized Glass Substrate
grand avb
errorc
contact angled
exptl
liquid surface
no. of
lisuids"
contact angle
limits (i)
difference
temp, "C
tension, mJ/m2
expts/drops
24
39.0
215
M
58.3
1.1
0.3
-2.7
24
39.0
215
MS
61.0

a-dSee footnotes of Table 111.

on the location of specific points of features of the drop shape,
and it is generally applicable to all axisymmetric liquid-fluid
interfaces; i.e., sessile or pendant drops and contact angles both
greater than or less than 90'. The technique has an uncertainty
of less than f0.4' (standard error of the mean for a sample size
of three). I t is also objective, being independent of the skill and
experience of the operator, and is relatively straightforward, both
in terms of the apparatus and the details of its use.
Drops are photographed in the horizontal plane and approximately 40 coordinate points are selected arbitrarily from the
profile utilizing either a manual digitizing tablet or a digital image
analyzer. This array of points is then fitted, in a least-squares
sense, to the Laplace equation which yields both the contact angle
and the liquid surface tension (for details see ref 8 or 9).
Apparatus and Procedure. Table I lists the measured surface
tension-temperature relation for each of the seven liquids. By
controlling the temperature of the contact angle experiment, it
was possible to match more exactly the total surface tensions of
the various liquids. All of the experiments were, therefore,
performed in a temperature-controlled chamber.
Advancing contact angles were produced by growing sessile
drops through a small hole in the center of the solid substrate
with a motorized syringe drive.g Care was taken to eliminate
vibrations in the system and to ensure that a given solid substrate
was always oriented in the same direction so that each drop was
photographed on the same portion of the surface.
The same substrate was used for each liquid in a given pair
(one "dispersive" and one "nondispersive" liquid). Prior to an
experiment, the test surfaces were soaked for 20 min each in two
changes of ethanol (absolute) and were then transferred to
methanol (Fisher 99.9 mol %) and sonicated for 15 min. After
two final rinses in methanol, the surfaces were placed in a small
(10) Rotenberg, Y.; Boruvka, L.; Neumann, A. W. J. Colloid. Interface
Sci. 1983, 93, 169.
(11) Boyce, J. F.; Schurch, S.;Rotenberg, Y.; Neumann, A. W. Colloids
Surf. 1984, 9,307.

desiccator and dried under vacuum by using a water aspirator.

The substrates were finally placed in a second desiccator and left
under vacuum for at least 1.5 h prior to use. Similar cleaning
procedures were employed for all syringes and needles used in
the handling of the liquid samples.
An experiment was begun by growing a drop on a substrate
disk and then taking a series of photographs of the meridional
profile. The drop volume was then increased by adding liquid
through the hole in the center of the disk and this larger drop
was in turn photographed. In this way a single experiment
consisted of photographing a series of successively larger drops
(usually three) on a given substrate. Individual drops were
photographed 3 times, refocusing the microscope/camera each
time, and the contact angles from these three photographs were
averaged to provide a mean contact angle for that particular drop.
The mean contact angle for a single experiment, usually consisting
of three differently sized drops, was simply the average of the
individual mean contact angles for each drop. In other words,
the contact angle reported for a single experiment, of one liquid
or a given substrate, was the average of (usually) nine ADSA
results, comprising three replications for three distinct drops. In
some cases, only two different drops were photographed in a given
experiment.

Results and Discussion

The results of the contact angle experiments are rep o r t e d in Tables 111, IV, and V for the solid substrates
Teflon FEP sample A, FEP sample B, and siliconized glass,
respectively. The data are grouped i n pairs according to
the matched surface tensions of the liquids used. The first
liquid i n each pair is the one which is completely or overwhelmingly composed of London dispersion forces, and
the second liquid has relatively large nondispersion forces.
The f o u r t h column i n these tables, the contact angle difference, is defined as the contact angle of the first liquid
minus that of the second liquid. In four of the eight cases,

Langmuir, Vol. 2, No. 5, 1986 623

Solid Surface Tension

this difference exceeds the range of the combined error
limits, indicating the possible influence of some systematic
factor, distinct from the random error which contributes
to the error limits. The average of the eight contact angle
differences is +0.4. The following is a brief discussion
of the possible explanations for these results, in terms of
both the theory of surface tension components and that
of the equation of state.
The Fowkes equation for solid-liquid interfacial tensions,
YSL

= 7s

+ YL - 2(YsdYLd)/2

is strictly applicable only to situations in which a t least

one phase is a saturated hydrocarbon (n-alkane, paraffin
wax, etc.) since this ensures that only dispersion forces are
operative within that phase. Here, ys and yL are, respectively, the solid and liquid surface tensions (neglecting,
as is customary, equilibrium spreading pressures); ysL is
the solid-liquid interfacial tension; and ysd and yLdare the
dispersion components of the solid and liquid surface
tension, respectively. The presence of the two different
dispersion components under the square root sign in eq
2 should not be construed as implying the ability to treat
the interaction of nondispersion forces in the two phases
in a similar fashion.
Equation 2 predicts that if ys and yL are fixed, then ysL
will vary inversely with yLd. With respect to the contact
angle experiments, the dispersive liquid in each pair
should, therefore, have the smaller contact angle. As was
mentioned above, however, the average contact angle
difference for the eight liquid pairs was +0.4O, indicating
that the opposite trend was more prevalent. On the average, the dispersive liquid had a slightly larger contact
angle than the more nondispersive liquid, in direct
contradiction with the predictions of eq 2. Considering
the four cases where the contact angle difference exceeds
the error limits, in two of these (dibenzylamine with benzaldehyde and with methyl salicylate on FEP A) the difference is positive (contrary to eq 2), while in the other two
cases it is indeed negative (pentadecane and heptaldehyde
on FEP B and 1-methylnaphthalene and methyl salicylate
on siliconized glass).
Equation 2 may be combined with the Young equation,
eq 1, to yield

+ COS e)

2(7Sd)1/2= (yL/yLd)/2yD1/2(1

(3)

For a given pair of liquids (denoted 1 and 2) on a single

substrate, the left-hand side of eq 3 is constant so that
-=--

yL:

YL:

YL*

YL1 YL1

YL*(

1 cos e 2

1 + cos

)2

(4)

el

Considering liquid 1to be the dispersive liquid (the first

liquid listed in each pair in Tables 111-V) and assuming
for these liquids that the dispersion fraction is that listed
in the last column of Table I1 (from the Beerbower correlation a t 25 C), eq 4 may be used to calculate the implied dispersion fraction of liquid 2 (the left-hand side
of eq 4). Note that eq 4 can also be used to give the ratio
of the dispersive fractions of liquids 1and 2, without
regard to the Beerbower correlation.
Equation 4 is applicable to situations where the solid
substrate is comprised only of London dispersion forces.
This condition is satisfied by the use of both Teflon FEP
and siliconized glass which has a surface consisting of
methyl groups and thus behaves as a saturated hydrocarbon.
Table VI presents the results of these calculations for
the nondispersive liquids on the three substrates. The last

Table VI. Fowkes Equation Predictions of the Fraction of

the Total Surface Tension due to Dispersion Forces for the
Nondispersive (2) Liquids in Each Pair

liquids

exptl
temp, OC
24
24
3
3
3
3
39
39
14
14

(Beerbower)
(Table 11)
FEP A
0.97
0.56
0.94
0.76
0.94
0.56
1.00
0.56
1.00
0.58

y~:/r~,

(eq 4)
0.97
1.02
1.00
0.98
1.01

FEP B
3

0.94
0.56
1.00
0.56

0.98

Siliconized Glass
24
0.97
24
0.56

0.92

39
39

1.01

Liquid symbols as in Table 111.

column in this table is the Fowkes-theory prediction of the

dispersive fraction of the 2 liquid (the nondispersive one)
within each liquid pair. In all cases, this dispersive fraction
is very close to 1.00, indicating that eq 2 predicts that the
2 liquids are just as dispersive as are the 1 liquids. This
is clearly contrary to the predictions of the Beerbower
correlation and, in general, to the expectations based on
solubility parameters, molecular structure, and molecular
properties. Within the context of the Fowkes approach,
it therefore needs to be asked why the apparently significant differences in the relative magnitudes of dispersion
forces have not manifested themselves in concomitant
differences in the observed contact angle measured with
each pair of liquids.
One possible answer is that the 2 liquids are, in effect,
almost completely dispersive in behavior with the nondispersion contributions being too small to affect the
contact angle noticeably. As was mentioned above, however, this is in considerable disagreement with the independent measures of the relative importance of nondispersion forces; viz., solubility parameters, the Beerbower
correlation, and dipole moments. Nevertheless, if the 2
liquids are indeed accepted as being almost completely
dispersive, then the conclusions remain significant within
the context of the Fowkes approach. If liquids as apparently nondispersive as the 2 liquids of Table VI are found
to behave as if they were overwhelmingly dispersive (as
are the 1 liquids), it in effect means that in many
practical situations it is unnecessary to evaluate the nondispersion components of surface tension; the wetting
behavior is well modeled on the basis of a single (dispersion) component of surface tension equal to the total
surface tension. In other words, the equation of state
approach is applicable.
Alternatively, if it is agreed that there is a significant
difference between the liquids with regard to the types and
relative magnitudes of the intermolecular forces, then it
must be concluded that eq 2 has failed to detect these
differences in terms of surface tension components.
Therefore, within the context of the theory of surface
tension components there are primarily two possible responses to the observed contact angles. The first may be
to repudiate the independent measures of a liquids

624 Langmuir, Vol. 2, No. 5, 1986

Spelt et al.

Table VII. Equilibrium Spreading Pressures Required To

Make the Contact Angles Equal in Each Liquid Pair
liquidb
exptl
Ad,
Ad,,
AB,,
Ke,e
pair
temp, C
deg
deg
deg
mJ/m2
FEP A
24
-0.2
0.0
-0.2
0.1
M+MS
+3.4
1.4
D+B
3
+2.0
-1.4
+0.4
+2.1
0.9
+2.5
D+MS
3
-1.0
-0.8
+0.2
0.3
39
P+H
P + EC
14
+0.6
0.0
+0.6
0.1
FEP B

D + MS
P+H

+3.1

39

-1.2

M+MS

39

+0.4
+0.2

+2.7
-1.4

1.2
0.4

Siliconized Glass
-2.7
0.0

-2.7

1.0

Calculations based on the Neumann equation of state. Liquid

symbols as in Table 111. The first liquid in each pair is the 1
(dispersive) liquid and the second is the 2 liquid. A6 = d1 - d2;
Ad,, contact angle difference due to the difference between y~~and
yL2; Adr, contact angle difference due to
K,, equilibrium
spreading pressure responsible for Adr.
dispersivecharacter such as solubility parameters, dipole
moments, and molecular structure. The argument would
then be that the liquids in each polar-nonpolar pair are
equally dispersive and hence the contact angles should
indeed be the same. The second possible response would
acknowledge the differences in the liquids within each pair
and would conclude that the Fowkes equation does not
correctly identify the anticipated nondispersion components of surface tension.
In contrast to the above dichotomy, the present contact
angle results are fully consistent with the equation of state
approach to interfacial tensions. Recall that since the total
liquid surface tensions are constant within a given pair of
liquids, the contact angles are predicted (by an equation
of state) to be equal on a single solid substrate. This does
appear to be largely the case, although explanations must
be found for the small, but finite, contact angle differences
which persist. There are several ways to account for these
discrepancies: (1) vapor adsorption leading to small
equilibrium spreading pressures; (2) liquid contamination;
(3) nonmaximal advancing contact angles.
With respect to the last of these points, dynamic contact
angle experiments12 have demonstrated that a static
advancingcontact angle may be as much as 3 O below the
true Young contact angle encountered at very low threephase line velocities.
Regarding the first possible explanation, although it is
customary to neglect vapor adsorption (and, hence,
equilibrium spreading pressures) in all practical contact
angle measurements, it is important to appreciate that this
is an approximation, albeit a good one, for most low-energy
s01ids.l~ It is widely accepted that surfaces such as Teflon
and polyethylene contain a small fraction of hydrophilic
sites which may lead to spreading pressures on powdered
Teflon of over 3 mJ/m2 with octane and approximately
2 mJ/m2 with water.13 The manner in which these powder
measurements relate to solid surfaces of Teflon is in
question, but it seems plausible to expect, under certain
circumstances, equilibrium spreading pressures of the order
of 1 mJ/m2.
Table VI1 provides an estimate of the equilibrium
spreading pressure required to make the contact angles
equal in each liquid pair. The observed contact angle
(12) Cain, J. B.; Fancis, D. W.; Venter, R. D.; Neumann, A. W. J.
Colloid Interface Sci. 1983, 94, 123.
(13) Good, R. J. A.C.S. Symp. Ser. 1975, No. 8.

difference, AO, between two liquids in a pair may be considered in two distinct parts, one due to the small differences in the liquid surface tension (AO,) and the other due
to vapor adsorption (Ae,). A0 is defined as O1 - O2 where
1and 2 denote respectively the dispersive and nondispersive liquids in a given pair.
In order to calculate AO, and then AOr it is necessary to
make use of an equation of state for interfacial tensions,
ysL = ysL(ysv,yLv). Although an explicit formulation of
such an equation has not yet been achieved analytically,
there do exist empirical equations of state, of which only
one has been widely used and t e ~ t e d . ~ ,This
~ J ~particular
equation, due to Neumann et al., is based on a wide variety
of contact angle data on low-energy surfaces, and in
practice, it is implemented as a computer program2(which
has been adapted to hand-held calculator~l~)
or in the form
of tables.16
The evaluation of AO, begins with the calculation of the
solid surface tensions of the three substrates using the
contact angle data in Tables 111-V. The least-squares
relations for ysv, from the Neumann equation of state, are
as follows: 19.3 - (0.0632)T mJ/m2 and 20.4 - (0.0722)T
mJ/m2 for FEP A and FEP B, respectively. Knowing the
solid surface tension at each experiment temperature, it
is possible to predict the contact angle difference (AO,) due
to the known difference in the measured liquid surface
tensions yL1- yL2where the subscripts have the same
meaning as above.
The contact angle diffference attributable to adsorption
is then given by AO, = A0 - AO,. If AO, is positive (the
dispersive, 1liquid is considered to be adsorbed) then
K, is estimated as follows: K, = ys, - ysvl, where ysv, and
ys, are calculated with the Neumann equation of state and
the contact angle-liquid surface tension data, 01,yLvland
61 - AOT, yLvl, respectively. Here, O1 and yLvl are the
contact angle and the surface tension, respectively, of the
dispersive liquid in each pair listed in Tables 111-V. An
analagous procedure, utilizing O2 and yLv2,is employed to
estimate K, if AO, is negative.
Of the eight hypothetical equilibrium spreading pressures listed in Table VII, six are less than or equal to 1
mJ/m2. As demonstrated by Good,13it is not unreasonable
to assume that such spreading pressure can occur on
surfaces of Teflon FEP and siliconized glass. It is,
therefore, concluded that the observed differences in the
contact angles within a given liquid pair may reasonably
be attributed to experimental error and specifically to
vapor adsorption. The present contact angle data are,
thus, seen to be consistent with the predictions of the
equation of state approach and provide an independent
experimental verification of this theory.
Finally, mention should be made of a second source of
independent experimental evidence which may be used to
evaluate the theory of surface tension components and that
of the equation of state. The rejection or inclusion of
microscopic particles at advancing solidification fronts may
be predicted thermodynamically and then verified by
direct experimental ob~ervation.~This type of experiment, thermodynamic prediction followed by direct observation, has consistently confirmed the accuracy of the
(14)Neumann, A. W.; Spelt, J. K.; Smith, R. P.; Francis, D. W.; Rotenberg, y.;Absolom, D. R. J. Colloid Interface Sci. 1984, 102,278.
(15) Taylor, C. P. S. J . Colloid Interface Sci. 1984, 100, 589.
(16) Neumann, A. W.; Absolom, D. R.; Francis, D. W.; van Oss, C. J.
Sep. Purif. Methods 1980, 9, 69.
(17) Neumann, A. W., Omenyi, S. N.; van Oss, C. J. J. Phys. Chem.
1982, 86, 1267.
(18) McClellan, A. L. Tables of Experimental Dipole Moments; W.
H. Freeman: San Franciso, 1963.

Langmuir 1986,2,625-630
Neumann equation of state.17 The application of eq 2, on
the other hand, is known to result in false predictions of
particle behavior in a number of ~ y s t e m s . ~ ~ ' ~
The findings reported here give rise to some regrets. The
idea of subdividing surface tensions into components which
arise from the action of the individual intermolecular forces
is very appealing. A completely successful theory of this
nature would be an immensely valuable link between
macroscopic and microscopic thinking. It should be understood that the experimental results presented here do
not preclude the existence of such a theory. While it has
become apparent that present day values are not dispersion components of surface tension, such components
might well exist and contribute to the overall surface
tension, although probably not in an additive fashion. We
see the Fowkes approach as a decisive first step in the
development of this area of surface science.
(19) Smith, R. P. Ph.D. Thesis, University of Toronto, 1984.

625

Conclusions
The contact angles of pairs of liquids on a single solid
substrate were found to be equal when the surface tensions
of the liquids were the same, regardless of the relative
magnitudes of the dispersion forces comprising each of the
two liquids. This observation is in direct conflict with the
theory of surface tension components. It is, however,
consistent with the expectations of the equation of state
approach, in which the solid-liquid interfacial tension is
only a function of the total solid and liquid surface tensions.
Acknowledgment. Supported in part by the Natural
Science and Engineering Research Council of Canada
(A8278), the Medical Research Council (MT 5462, M T
8024, MA 9144), and the Ontario Heart and Stroke
Foundation (4-12, AN-402). D.R.A. acknowledges gratefully support of the Ontario Heart Foundation through the
receipt of a Senior Fellowship.

Interactions in Concentrated Nonaqueous Polymer Latices

Ivana Markovie, R. H. Ottewill,* and Sylvia M. Underwood+
School of Chemistry, University of Bristol, Bristol B S 8 1 TS, U.K.

Th. F. Tadros
ICI Plant Protection Division, Jealott's Hill Research Station, Bracknell, Berkshire, U.K.
Received February 25,1986. I n Final Form: June 25, 1986
Particles of poly(methy1methacrylate) stabilized by poly(l2-hydroxystearicacid) can be used to form
dispersions of high volume fraction in dodecane. In the present work volume fractions between 0.23 and
0.42 were examined by small-angleneutron scattering. From these measurements both the structure factor,
S(Q),and the pair correlation function,g(r), were derived. By use of a model interaction potential function,
based on the mean spherical approximation, it was found that the particle-particle interaction was soft
at low volume fractions and became harder at high volume fractions, indicating that in the latter condition
some compression of the stabilizing layer occurred.

Introduction
Following the introduction of the concept of dispersion
polymerization in nonaqueous media1g2methods have been
developed for the preparation of stable monodisperse
polymer latices in hydrocarbon^.^-' These have proved
useful model colloidal dispersions for fundamental studies
in media of low dielectric constant. A system of this type,
which we have used in previous s t ~ d i e s ~is, ~composed
,~*~
of a core particle of poly(methy1 methacrylate) with an
outer layer of a "comb" polymer, poly( 12-hydroxystearic
acid), covalently attached to the core particle so that the
oleophilic "teeth" project into the dispersion medium. In
hydrocarbon media such as dodecane it seems reasonable
to conclude that electrostatic effects are absent and hence
the particles are sterically stabilized. Consequently, the
range of interaction forces between these particles would
not be expected to be significant beyond a range of about
twice the length of the stabilizer chains,1 in the present
case the "teeth" of the comb polymer.
Although a significant number of theories have been
proposed to account for the phenomena of steric stabili-

* To whom correspondence should be addressed.

t Present

address: Dulux Australia Limited, Clayton, Victoria,

Australia.

0743-7463/86/2402-0625$01.50/0

zation, and these have been well surveyed by Napper,"

there is still a sparsity of direct experimental information
on the nature of the interaction potentials between sterically stabilized particles. However, our previous preliminary study on this type of system using small-angle neutron scattering4 and subsequent theoretical developm e n t ~ ' ~has
- ~ indicated
~
that this technique provides a
(I) Osmond, D. W. J.; Walbridge, D. J. J. Polym. Sci. 1970, (230,381.

(2) Barrett, K. E. J. Dispersion Polymerization I n Organic Media;

Wiley: London, 1975.
(3) Cairns, R. J. R.; Ottewill, R. H.; Osmond, D. W. J.; Wagstaff, I. J.
Colloid Znterface Sci. 1976,54, 45.
(4) Cebula, D. J.; Goodwin, J. W.; Ottewill, R. H.; Jenkin, G.; Tabony,
J. Colloid Polym. Sci. 1983,261,555.
(5) Antl, L.; Goodwin, J. W.; Hill, R. D.; Ottewill, R. H.; Owens, S. M.;
Papworth, S.;Waters, J. A. Colloids Surf. 1986, 17,67.
(6) Dawkins, J. V.; Taylor, G. Colloid Polym. Sci. 1980,258, 79.
(7) Everett, D. H.; Stageman, J. F. Faraday Discuss. Chem. SOC.1978,

--.(8) Cairns, R. J. R.; van Megen, W.; Ottewill,R. H. J.Colloid Interface

65. 2.70.

Sci. 1981, 79, 511.

(9) van Megen, W.; Ottewill, R. H.; Owens, S. M.; Pusey, P. N. J.

Chem. Phys. 1985,82, 508.
(10)Koelmans, H.: Overbeek, J. Th. G. Discuss. Faraday SOC.1954.
18,52.
(11) Napper, D. H. Polymeric Stabilization of Colloidal Dispersions;

Academic Press: London, 1983.

(12) Hayter, J. B.; Penfold, J. Mol. Phys. 1981, 42, 109.
0 1986 American Chemical Society